Reasearch On Idengtification Model For Design Defects Of Cylindrical Gear Load Capacity

Design defects identification of cylindrical gear is key factor for ensuring gear design quality, reducing gear design cycle, improving gear security and reliability. On the basis of researching the root of design defects of cylindrical gear load capacity, identification for design defects of cylindrical gear strength, design defects of cylindrical gear stiffness and design defects of cylindrical gear installation error are studied deeply and systematically in this thesis.Cylindrical gear strength is an important indicator of gear load capacity. Due to large numbers of complex constraints in the gear strength design, it’s difficult for traditional design method to ensure that all of constraints are simultaneously satisfied by design variables. Utilizing nonlinear mapping ability of BP neutral network, reflecting the physical relationship between design variables of cylindrical gear and structural properties as mathematic relationship between neutral network input and output, identification model for design defects of cylindrical gear strength based on BP neutral network are established. Utilizing parallel processing ability for complex problem of genetic algorithm, optimizing globally for BP neutral network model after training, design variable space satisfying design requirements of gear strength is obtained, design defects of cylindrical gear strength are corrected optimally.An identification method for design defects of cylindrical gear stiffness based on combination of improved energy method and finite element method is proposed. Firstly shear stiffness and the matrix deformation stiffness are introduced into traditional energy method. Time-varying meshing stiffness of cylindrical is calculated by improved energy method. Secondly the Fourier decomposition is used to change the type of time-varying meshing stiffness. In the new type of stiffness formula, stiffness excitation is composed by a series of sine wave. Finite element software is used to analyzing impacts of periodic stiffness excitation on gear stress distribution. Stiffness design defects of original design could be confirmed by whether the static strength and fatigue strength meet design requirements. At last, correcting optimally gear design parameters, gear stress concentration and gear alternating stress in meshing process is reduced resulted by reduced time-varying meshing stiffness of gear, design defects of cylindrical gear stiffness is corrected optimally.Oversized gear installation error is the important cause of gear failure such as breakage, tooth wear, tooth surface pitting, which cause severe shock and vibration in gear meshing process. According to involute gear geometry and modern theories of gear meshing, tooth surface equation of involute cylindrical gear is derived, and tooth contact analysis model containing gear installation error is established. Solving the tooth contact analysis model, gear tooth contact trace and contact ellipse derived. It shows that installation error have influenced on position and area of contact ellipse. Installation error design defect of original design could be confirmed by whether the position and area of contact ellipse meet design requirements. Taking the domain function and area function of gear tooth contact ellipse as objective function, the gear installation error as design variables, the tolerance limit of gear installation error is obtained. Adjusting installation error into tolerance limit, design defect of cylindrical gear installation error is corrected.Identification model for design defects of cylindrical gear strength, identification model for design defects of cylindrical gear stiffness and identification model for design defects of cylindrical gear installation error provide theoretical basis for design defect identification of cylindrical gear load capacity.